PrismHub
Start building
CLEAR INPUTS. INFORMED DECISIONS.

Pay for what you use.
Plan for what you build.

Model your costs before you commit. Understand the inputs behind a token-based workload, from a single request to a month of conversations.

Usage-based

Plan around the input and output tokens your workload consumes.

Model-specific

Choose the model that fits your task and check its applicable rates.

Understandable

Separate request volume, token length, and rates in your estimates.

OpenRouter reference rates + 5%.Input, output and cache rates are synced together. Trial credits are allocated by your administrator.
WORKLOAD COST CALCULATORCurrent PrismHub rates

Put a number on your idea.

Choose a model. Estimate fresh input, cache usage and output separately.

Input
Unavailable
Output
Unavailable
Cache read
Unavailable

Cache reads and writes are subsets of total input, never additional input. Estimates use current rates; actual requests freeze their rates at start. Text tokens only.

USD / 1M tokens · OpenRouter reference rates + 5%Reference pricing is unavailable. This model cannot be called.
ESTIMATED MONTHLY COST—

Choose a priced model and valid quantities. Cache reads + writes must not exceed total input.

Input
—
Output
—
Cache read
—
Cache write
—
Cost per request
—
Σ (category tokens × PrismHub rate)
× request count
How to budget for tokens →
LIVE MODEL PRICING

One rate card. Every token accounted for.

OpenRouter reference rates + 5%. Input, output and cache usage are priced separately.

Loading current rates…

Pricing, explained.

YOUR NEXT CHAPTER

Build without boundaries.

Find your model. Connect your stack. Make something that matters.

Start building